Intrusion detection model based on frequent pattern mining over data streams
XU Ying-mei · Journal of Shaanxi University of Technology · 2011
Computer network intrusion typically has the characteristics of high frequency,Therefore,the data stream mining of repeating elements and given the frequency index are the important basis for visit to identify whether it is normal.An advanced method(AFP) for mining the frequent patterns of data streams is proposed.AFP tree algorithm uses the sliding window technique.As data stream flows,the contents of the data stream are captured with a compact prefix-tree by scanning the stream only once.And the algorithm is used in the intrusion detection model,the normal data and abnormal data mining online.The conflicts between limited memory and unlimited data flow are solved.The experimental results show that the model has a high alarming rate and a lower false rate.